At the most elemental level, the central nervous system is involved in processing information represented by temporal patterns of action potentials in individual nerve cells. The olfactory system is an ideal brain region in which to study how sensory information is encoded. While olfactory transduction occurs in specialized olfactory receptor neurons, the actual work of generating temporally-modulated odorant-specific neural discharges begins in the olfactory bulb. Here the simple monotonic input from receptor neurons activates spatially-defined subpopulations of output neurons (mitral cells). However, mitral cells do not simply relay sensory information on to third-order brain areas. Instead, receptor neuron input interacts with unusual intrinsic currents in mitral cells and dendrodendritic inhibitory synaptic connections that mediate the strong lateral interactions between different subpopulations of mitral cells. This proposal examines the cellular mechanisms that underlie both the unusual intrinsic properties of olfactory bulb neurons and the dendrodendritic inhibitory circuits that underlie recurrent and lateral inhibition. We employ a combination of intracellular recording and neuropharmacological tools to investigate why recurrent inhibition in the olfactory bulb appears to be dependent upon NMDA receptors and how it is modulated by cholinergic receptors. We also have discovered a new class of interneurons in the olfactory bulb that become persistently active following transient stimuli. We propose experiments to define the intracellular signaling mechanisms that mediate and modulate persistent activity. Finally, we propose a series of experiments to determine how mitral cells integrate these different synaptic and intrinsic currents when generating physiological discharge patterns. Understanding how sensory information is represented as spatio-temporal discharge patterns will have wide ranging significance beyond the immediate goal of understanding the synaptic organization of the olfactory bulb. Electrical stimulation of different CNS regions has been shown to be therapeutic in neurological disease. Currently, these interventions are based on empirical findings, often employing non-physiological tetanic stimulus trains, rather than patterned stimuli at physiological frequencies. One (1) outcome from the present study is likely to be a better understanding of how physiological patterns of activity are generated by local circuits. Our work may lead to new therapeutic strategies for treating neurological diseases such as Parkinson's disease and epilepsy that employ biologically-inspired patterned stimulus trains.